Graduate Student Spotlight: Marzia Oceno, PhD Candidate, Political Science

“Will Any Woman Do? Feminists’ and Non-Feminists’ Support for Female Political Candidates”

The Boyd/Williams Dissertation Grant for Research on Women and Work is awarded annually to PhD students whose dissertations promote knowledge about and enhance understanding of the complexities of women’s roles in relation to their paid and unpaid labor, including volunteerism, domestic work, and political activity. In 2018, IRWG awarded two students the Boyd/Williams Grant: Sandhya Narayanan and Marzia Oceno.

Marzia Oceno studies trends in voting patterns in American politics. One trend is that people tend to support political candidates who come from their same social identity group. However, the literature on descriptive representation presents very mixed results when it comes to gender. In fact, most previous studies find no overwhelming gap in support for female candidates by female voters.

“We keep hearing about the under-representation of women in politics at all levels of government,” says Oceno. “The puzzle is that even when a lot of women candidates run, there is not consistent support for them by women.”

The 2016 presidential election compelled Oceno to solve this puzzle. It wasn’t Hillary Clinton’s loss that was most curious, rather, it was the breakdown of support for Clinton by women voters. According to exit polls, 54% of female voters supported Clinton in 2016. However, a similar percentage – 55% – of women turned out for Barack Obama in 2012. She wondered, “Why was shared gender between voters and candidates such a weak predictor of voting behavior in 2016?”

These results also overshadow subgroup divides: 94% of black women and 69% of Latino women voted for Clinton, while only 43% of white women did so. “Basically, when I saw that,” explains Oceno, “I decided to focus on white women’s voting patterns, because white women are very divided.” Oceno suspected that the inconsistent findings related to the impact of gender identification on electoral behavior stem from the neglect of two highly salient identities at the subgroup level: feminism and non-feminism.

Using funds from her Boyd/Williams Grant, she conducted an online survey experiment to examine how women’s feminist or non-feminist identity affects evaluations of female political candidates. The sample included 551 U.S. adult women. Participants were randomly assigned to read about a fictitious, upcoming nonpartisan mayoral election featuring a female candidate in Davenport, Iowa. The candidate was described as a moderate without explicitly mentioning any partisan affiliation. The experiment manipulated two factors: a female candidate’s gender subgroup identity (either feminist or non-feminist) and her race (white or black).

Oceno’s experiment confirmed her suspicions. While there were no statistically significant differences in levels of support based on the candidate’s race, an important pattern emerged with regard to the feminist and non-feminist conditions. Subjects who self-identified as feminists felt significantly warmer toward both white and black feminist candidates, viewed them more favorably, were more likely to donate money for their campaigns, and were more likely to vote for them, regardless of race.

On the flip side, subjects who self-identified as non-feminists were not necessarily more supportive of the non-feminist candidates, but they were strongly opposed to the feminist candidates. She explains: feminist self identification and the resulting preference for feminist candidates, stems from ingroup positivity. In the opposite direction -- an example of outgroup negativity -- subjects who identify as non-feminists are not necessarily more supportive of the non-feminist candidates, but they are strongly opposed to the feminist candidates. This aversive reaction to feminist candidates could (at least partially) explain Clinton’s lack of support from white women voters in 2016.

In her experiment, Oceno never explicitly used the feminist label to describe her fictitious candidates. The candidate’s categorization with either the feminist or non-feminist subgroup was conveyed through her personal biography and her policy platform. Even so, non-feminist subjects gave significantly more negative evaluations of feminist candidates. “When people use the [feminist] label, you can imagine the results to be even more dramatic,” says Oceno.

Looking ahead, Oceno will pursue more data collection. First, she will conduct a survey to delve more deeply into individuals’ conceptions of feminism and non-feminism. The second study will examine whether the female electorate punishes women candidates (maybe more than politically identical men) for adopting the feminist label.

She hopes that examining feminism and non-feminism as lenses through which individuals interpret the political spectrum will lead to a better understanding of how gender subgroup identities drive public opinion and political outcomes.